[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M77JEBCAN2YMDW0N0FSFVVCK":3,"skill-same:01M77JEBCAN2YMDW0N0FSFVVCK":43},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":23,"commitSha":24,"skillDirectory":25,"forksCount":13,"repositoryHtmlUrl":26,"repositoryHomepageUrl":6,"ownerHtmlUrl":27,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":28,"install":29,"resources":42},"01M77JEBCAN2YMDW0N0FSFVVCK","manage-settings",null,"Any explicit Muse Code setting question or change (model, reasoning effort, \u002Fsettings) requires a silent read_skill call for bundled:manage-settings as FIRST ACTION—no assistant text or other tool first. Never use for repository, app, eval, or other non-Muse configuration.","raaOS\u002Fsystem_prompts_leaks","raaOS","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F214067397?v=4","01M77HV6K4ZTA4BZAGX00GTW6J","01M77HV6KX55HVRY679V25JXF8",0,"CC0-1.0","https:\u002F\u002Fgithub.com\u002FraaOS\u002Fsystem_prompts_leaks\u002Fblob\u002Fmain\u002FMeta\u002Fmuse-code\u002Fskills\u002Fmanage-settings\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: manage-settings\ndescription: Any explicit Muse Code setting question or change (model, reasoning effort, \u002Fsettings) requires a silent read_skill call for bundled:manage-settings as FIRST ACTION—no assistant text or other tool first. Never use for repository, app, eval, or other non-Muse configuration.\n---\n\n# Manage Settings\n\nExplain saved settings and defaults from the workspace's configuration, keep\nthem distinct from live\u002Feffective state, and make small settings edits only when\nthe user explicitly asks for that edit.\n\n## Scope\n\n- Manage Settings is exclusively for Muse Code-owned settings.\n- Never read, name, explain, or modify another agent's config, rules, skills,\n  home directories, or compatibility environment.\n- Use this skill for questions about product settings, config paths, reminder\n  toggles, TUI settings, provider\u002Fmodel settings, and tool or policy settings.\n- \"Product settings\" here means Muse Code itself. Repository settings,\n  application settings screens, library configuration, and eval-task config\n  work stay in the ordinary task workflow and MUST NOT load this skill.\n- Prefer the current workspace and current process environment over general\n  defaults.\n- Do not create issues, branches, commits, PRs, plugins, or skills.\n- Do not install, enable, disable, trust, activate, or run plugins or skills\n  unless the user explicitly asks for that action.\n- Do not print secrets. If a setting might contain a token, key, credential, or\n  opaque auth value, describe whether it is present without revealing the value.\n\n## Decide Before Any Tool Call\n\n- Manage Settings owns persistent saved settings, not current runtime state.\n- An explicit request to change a setting is a saved-settings edit request; the\n  user does not need to say \"saved default\" or name `settings.json`.\n- For questions about controls, supported tiers, or timing, answer from the\n  stable contract below. Read config when the user asks for a saved value or a\n  persistent setting change.\n- The first settings-file CONTENT access MUST be `read_file` on the resolved\n  active path. When the absolute path is not already known, one read-only Bash\n  call may resolve it from the environment, but that call may print only the\n  resulting path and MUST NOT touch the filesystem. Use this shape without\n  adding `ls`, `test`, `stat`, `cat`, or another command:\n  `python3 -c 'import os; print(os.path.join(os.environ.get(\"XDG_CONFIG_HOME\") or os.path.join(os.environ[\"HOME\"], \".config\"), \"muse\", \"settings.json\"))'`.\n  YOLO mode alone never authorizes direct Bash file access.\n- Use `edit_file` or `write_file` for the smallest safe JSON change. If those\n  tools reject the same app-private config path, only when the startup security\n  context explicitly says YOLO mode is already active (approval bypassed,\n  shell sandbox off, and workspace trusted), a structured shell JSON edit may\n  be used after the file tools reject that exact path in this turn. Use a JSON\n  parser plus atomic replace and reread; never use text substitution.\n- A shell file-access fallback must operate on that one resolved path. It MUST NOT test a\n  second root, branch on file existence, use `$HOME` after resolving an\n  `XDG_CONFIG_HOME` path, list the directory, or `cat` the whole settings file.\n  Emit only the requested safe key values and the minimal preservation evidence.\n- Never ask the user to change security mode just to edit settings.\n- Never recommend disabling sandbox, approvals, or another security control to\n  inspect or change settings.\n- Renderer settings such as reasoning summaries and verbose output cannot be\n  simulated in the assistant response. Explain the owning control instead of\n  promising equivalent behavior.\n\n## Read Before Answering\n\nFirst classify the request. Questions about product controls, supported tiers,\nor timing use the stable contract below and do not require a settings-file read.\nRead the file for a saved-value question or an explicit setting change. A\nrequest to change a setting means a persistent file edit; it does not authorize\na claim about current runtime state.\n\nBefore explaining or editing a saved value:\n\n1. Identify the settings file. Prefer `$XDG_CONFIG_HOME\u002Fmuse\u002Fsettings.json`\n   when `XDG_CONFIG_HOME` is set; otherwise use `$HOME\u002F.config\u002Fmuse\u002Fsettings.json`.\n2. Read that file when it exists. After the optional path-only resolver, the\n   first content access MUST use `read_file`. If it rejects the app-private path\n   and startup evidence already says YOLO mode is active, the structured shell\n   JSON fallback above may read the same path.\n3. A missing file is the normal first-write state. For a read-only question,\n   say which path was checked, explain the relevant default, report that no file\n   changed, and stop. For an explicit unambiguous change, create the smallest\n   valid JSON document with `schema_version: 1` and only the requested setting;\n   the old value is absent plus its documented default. Do not create a file for\n   an ambiguous request.\n4. If the allowed read path is unavailable or the read fails for any reason\n   other than not found (permission denied, sandbox block, malformed JSON, I\u002FO\n   error), say which path was checked and which failure happened, explain the\n   relevant default behavior instead of inventing a saved value, and stop — do\n   not hunt for substitute configuration files, probe writability, or try a\n   different config root.\n5. Tie the answer to the concrete key path, such as\n   `runtime_capabilities[\"plugin:tbh-reminders:reminder:skill-reminder\"].enabled`.\n\n## Where Muse Code Keeps Its Settings\n\nThis skill answers from the settings surface only:\n\n- Config root: `$XDG_CONFIG_HOME\u002Fmuse`, else `$HOME\u002F.config\u002Fmuse` — holds\n  `settings.json` (saved settings and defaults), `auth.json`, and `trust.json`; report\n  auth\u002Ftrust presence only, never contents.\n\nThat config root is the whole answer surface for this skill. A configuration\nfile outside it is not Muse Code configuration — do not hunt for substitute\nconfiguration files, and never present an unrelated file as this session's\nactive or effective configuration, even when the Muse settings file is\nunreadable.\n\nResolve exactly one root. When `XDG_CONFIG_HOME` is set, `$HOME\u002F.config\u002Fmuse` is\noutside the active root: never read, list, test-write, or report it as a fallback\nor schema source.\n\nSession state, logs, crashes, and what-happened questions are not settings\nquestions: hand those to the `doctor` skill instead of reading data-dir\nfiles from here.\n\n## Saved Versus Live State\n\nA `settings.json` read-back proves saved configuration only; it never proves\ncurrent TUI, in-flight run, active tool surface, or provider-request state. A\nverified config write changes the permanent saved setting only. It is guaranteed\nto be loaded on the next Muse Code launch; do not claim that the current process\nreloaded it. Relaunch Muse Code, not the terminal application, when the user\nwants the saved value to become the new runtime default.\n\nOrdinary conversation does not invoke a TUI settings control. In the interactive\nTUI, `\u002Fmodels`, `\u002Feffort`, and `\u002Fsettings` are user-side controls; never claim\nthat you executed one. Telling the assistant `\u002Feffort` or describing a desired\nsetting in chat does not run that control.\n\nCurrent-runtime controls are outside this skill's mutation scope. `\u002Fmodels`,\n`\u002Feffort`, and `\u002Fsettings` are user-side controls; mention the relevant control\nwhen the user also wants an immediate current-session change, but never claim\nthat you executed it. For Meta, the persistent effort tiers are `minimal`,\n`low`, `medium`, `high`, `xhigh`, `max`, and `ultra`. `high` is the default Meta\nbaseline; `xhigh` is the opt-in premium precision tier. `ultra` remains the\nsaved client selection, uses `max` reasoning on the Meta wire (ADR 19425 D63), and currently\nenables proactive workflow\u002Fdelegation guidance when that tool surface is\navailable. It may proactively run multi-agent workflows and increase token\nusage quickly. Do not claim the proposed 64-slot unconfigured-root default is\ncurrent; that capacity change is still a target.\n\nYou may honor a behavioral preference for the current task without claiming a\nproduct setting changed only when it actually governs your response or tool\nchoices. For example, answer more briefly or avoid subagents when asked, while\nleaving the corresponding product setting untouched. Do not pretend to emulate\nrenderer behavior such as reasoning summaries or verbose-output repainting, or\nprovider, permission, or other external-setting behavior. Do not invent\nreasoning-effort tiers, budgets, ratios, mappings, or effects; use only the\nstable facts above. Do not infer token budget, quality, speed, or generic\nworkflow effects from tier names. In particular, do not say a higher provider\ntier performs more, deeper, longer, or better reasoning. The only current\nworkflow delta named by this contract is `ultra`'s proactive guidance; it does\nnot prove that provider reasoning itself is deeper or better.\n\n## Persistent Writes On Explicit Change\n\nDo not claim that a saved-file edit changed the current interactive TUI. An\nexplicit request such as \"use xhigh\", \"turn summaries off\", or \"set verbose\noutput to less\" authorizes the corresponding persistent setting edit when the\nrequest is unambiguous. If the requested value is ambiguous, ask which saved\nvalue to use and STOP without a mutation. A clarification timeout,\nauto-cancellation, dismissal, empty reply, or missing reply is not authorization:\nleave the setting unchanged. Never pick a value because it seems more common,\nlikely, stronger, or closer to a default.\n\nUse these stable key contracts:\n\n- `reasoning_effort` is the top-level persistent effort tier. Meta accepts\n  `minimal`, `low`, `medium`, `high`, `xhigh`, `max`, or `ultra`.\n- `tui.reasoning_summaries` is a persistent boolean; absent defaults to `true`.\n- `tui.verbose_output` is one of `less`, `edits`, or `more`; absent defaults to\n  `edits` (shown as `edits & writes`).\n- `tui.terminal_background` is one of `auto`, `light`, or `dark`; absent\n  defaults to `auto` (detect from the startup probe). `light`\u002F`dark` override\n  the detected terminal background for theme resolution at the next startup.\n\nWhen a write is requested:\n\n1. Read the current settings first.\n2. Make the smallest JSON change that satisfies the request. If the read proves\n   the file is missing, create a minimal object containing `schema_version: 1`\n   and only the requested setting. A denied or malformed file is not a missing\n   file and MUST NOT be replaced.\n3. Preserve unrelated keys, formatting-sensitive values, and unknown fields.\n4. Re-read the file to prove the saved value changed. Do not treat the reread as\n   live\u002Feffective-state evidence.\n5. Report the old saved value (or absent plus its effective default), the new\n   saved value, that it is permanent, that the current session is unchanged,\n   and that it takes effect on the next Muse Code launch. Relaunch Muse Code; do\n   not tell the user to restart the terminal application.\n\nFor a YOLO shell fallback, an earlier successful shell read does not replace the\nrequired `edit_file` or `write_file` attempt. Try the file mutation tool on the\nsame path first; use the atomic structured shell mutation only after that tool's\nexplicit denial in this turn.\n\nIf the allowed access path above still cannot read or write the active config,\nreport that no saved setting changed and stop. Do not probe another config root\nor bypass a security control.\n\n## Completion Report\n\nFor read-only answers, include:\n\n- the settings path checked;\n- the key path or default that answers the question;\n- whether any file was changed.\n\nFor a verified write, use this user-facing summary shape:\n\n```text\nSaved changes\n- \u003Ckey>: \u003Cold saved value or absent + default> -> \u003Cnew saved value>\n\nPermanent: yes\nCurrent session: unchanged\nEffective: next Muse Code launch\nReload: required - relaunch Muse Code; no terminal-application restart\nVerified: `settings.json` reread at \u003Cpath>\n```\n\nFor a setting such as `reasoning_effort` with higher-precedence startup input,\nappend `unless a launch flag overrides the saved value` to the\nEffective line. Name any setting intentionally left unchanged.\n\nIf the read, write, or verification fails, use:\n\n```text\nChanged: no\nReason: \u003Cexact path and failure>\nReload: not applicable\n```\n","# Manage Settings\n\nExplain saved settings and defaults from the workspace's configuration, keep\nthem distinct from live\u002Feffective state, and make small settings edits only when\nthe user explicitly asks for that edit.\n\n## Scope\n\n- Manage Settings is exclusively for Muse Code-owned settings.\n- Never read, name, explain, or modify another agent's config, rules, skills,\n  home directories, or compatibility environment.\n- Use this skill for questions about product settings, config paths, reminder\n  toggles, TUI settings, provider\u002Fmodel settings, and tool or policy settings.\n- \"Product settings\" here means Muse Code itself. Repository settings,\n  application settings screens, library configuration, and eval-task config\n  work stay in the ordinary task workflow and MUST NOT load this skill.\n- Prefer the current workspace and current process environment over general\n  defaults.\n- Do not create issues, branches, commits, PRs, plugins, or skills.\n- Do not install, enable, disable, trust, activate, or run plugins or skills\n  unless the user explicitly asks for that action.\n- Do not print secrets. If a setting might contain a token, key, credential, or\n  opaque auth value, describe whether it is present without revealing the value.\n\n## Decide Before Any Tool Call\n\n- Manage Settings owns persistent saved settings, not current runtime state.\n- An explicit request to change a setting is a saved-settings edit request; the\n  user does not need to say \"saved default\" or name `settings.json`.\n- For questions about controls, supported tiers, or timing, answer from the\n  stable contract below. Read config when the user asks for a saved value or a\n  persistent setting change.\n- The first settings-file CONTENT access MUST be `read_file` on the resolved\n  active path. When the absolute path is not already known, one read-only Bash\n  call may resolve it from the environment, but that call may print only the\n  resulting path and MUST NOT touch the filesystem. Use this shape without\n  adding `ls`, `test`, `stat`, `cat`, or another command:\n  `python3 -c 'import os; print(os.path.join(os.environ.get(\"XDG_CONFIG_HOME\") or os.path.join(os.environ[\"HOME\"], \".config\"), \"muse\", \"settings.json\"))'`.\n  YOLO mode alone never authorizes direct Bash file access.\n- Use `edit_file` or `write_file` for the smallest safe JSON change. If those\n  tools reject the same app-private config path, only when the startup security\n  context explicitly says YOLO mode is already active (approval bypassed,\n  shell sandbox off, and workspace trusted), a structured shell JSON edit may\n  be used after the file tools reject that exact path in this turn. Use a JSON\n  parser plus atomic replace and reread; never use text substitution.\n- A shell file-access fallback must operate on that one resolved path. It MUST NOT test a\n  second root, branch on file existence, use `$HOME` after resolving an\n  `XDG_CONFIG_HOME` path, list the directory, or `cat` the whole settings file.\n  Emit only the requested safe key values and the minimal preservation evidence.\n- Never ask the user to change security mode just to edit settings.\n- Never recommend disabling sandbox, approvals, or another security control to\n  inspect or change settings.\n- Renderer settings such as reasoning summaries and verbose output cannot be\n  simulated in the assistant response. Explain the owning control instead of\n  promising equivalent behavior.\n\n## Read Before Answering\n\nFirst classify the request. Questions about product controls, supported tiers,\nor timing use the stable contract below and do not require a settings-file read.\nRead the file for a saved-value question or an explicit setting change. A\nrequest to change a setting means a persistent file edit; it does not authorize\na claim about current runtime state.\n\nBefore explaining or editing a saved value:\n\n1. Identify the settings file. Prefer `$XDG_CONFIG_HOME\u002Fmuse\u002Fsettings.json`\n   when `XDG_CONFIG_HOME` is set; otherwise use `$HOME\u002F.config\u002Fmuse\u002Fsettings.json`.\n2. Read that file when it exists. After the optional path-only resolver, the\n   first content access MUST use `read_file`. If it rejects the app-private path\n   and startup evidence already says YOLO mode is active, the structured shell\n   JSON fallback above may read the same path.\n3. A missing file is the normal first-write state. For a read-only question,\n   say which path was checked, explain the relevant default, report that no file\n   changed, and stop. For an explicit unambiguous change, create the smallest\n   valid JSON document with `schema_version: 1` and only the requested setting;\n   the old value is absent plus its documented default. Do not create a file for\n   an ambiguous request.\n4. If the allowed read path is unavailable or the read fails for any reason\n   other than not found (permission denied, sandbox block, malformed JSON, I\u002FO\n   error), say which path was checked and which failure happened, explain the\n   relevant default behavior instead of inventing a saved value, and stop — do\n   not hunt for substitute configuration files, probe writability, or try a\n   different config root.\n5. Tie the answer to the concrete key path, such as\n   `runtime_capabilities[\"plugin:tbh-reminders:reminder:skill-reminder\"].enabled`.\n\n## Where Muse Code Keeps Its Settings\n\nThis skill answers from the settings surface only:\n\n- Config root: `$XDG_CONFIG_HOME\u002Fmuse`, else `$HOME\u002F.config\u002Fmuse` — holds\n  `settings.json` (saved settings and defaults), `auth.json`, and `trust.json`; report\n  auth\u002Ftrust presence only, never contents.\n\nThat config root is the whole answer surface for this skill. A configuration\nfile outside it is not Muse Code configuration — do not hunt for substitute\nconfiguration files, and never present an unrelated file as this session's\nactive or effective configuration, even when the Muse settings file is\nunreadable.\n\nResolve exactly one root. When `XDG_CONFIG_HOME` is set, `$HOME\u002F.config\u002Fmuse` is\noutside the active root: never read, list, test-write, or report it as a fallback\nor schema source.\n\nSession state, logs, crashes, and what-happened questions are not settings\nquestions: hand those to the `doctor` skill instead of reading data-dir\nfiles from here.\n\n## Saved Versus Live State\n\nA `settings.json` read-back proves saved configuration only; it never proves\ncurrent TUI, in-flight run, active tool surface, or provider-request state. A\nverified config write changes the permanent saved setting only. It is guaranteed\nto be loaded on the next Muse Code launch; do not claim that the current process\nreloaded it. Relaunch Muse Code, not the terminal application, when the user\nwants the saved value to become the new runtime default.\n\nOrdinary conversation does not invoke a TUI settings control. In the interactive\nTUI, `\u002Fmodels`, `\u002Feffort`, and `\u002Fsettings` are user-side controls; never claim\nthat you executed one. Telling the assistant `\u002Feffort` or describing a desired\nsetting in chat does not run that control.\n\nCurrent-runtime controls are outside this skill's mutation scope. `\u002Fmodels`,\n`\u002Feffort`, and `\u002Fsettings` are user-side controls; mention the relevant control\nwhen the user also wants an immediate current-session change, but never claim\nthat you executed it. For Meta, the persistent effort tiers are `minimal`,\n`low`, `medium`, `high`, `xhigh`, `max`, and `ultra`. `high` is the default Meta\nbaseline; `xhigh` is the opt-in premium precision tier. `ultra` remains the\nsaved client selection, uses `max` reasoning on the Meta wire (ADR 19425 D63), and currently\nenables proactive workflow\u002Fdelegation guidance when that tool surface is\navailable. It may proactively run multi-agent workflows and increase token\nusage quickly. Do not claim the proposed 64-slot unconfigured-root default is\ncurrent; that capacity change is still a target.\n\nYou may honor a behavioral preference for the current task without claiming a\nproduct setting changed only when it actually governs your response or tool\nchoices. For example, answer more briefly or avoid subagents when asked, while\nleaving the corresponding product setting untouched. Do not pretend to emulate\nrenderer behavior such as reasoning summaries or verbose-output repainting, or\nprovider, permission, or other external-setting behavior. Do not invent\nreasoning-effort tiers, budgets, ratios, mappings, or effects; use only the\nstable facts above. Do not infer token budget, quality, speed, or generic\nworkflow effects from tier names. In particular, do not say a higher provider\ntier performs more, deeper, longer, or better reasoning. The only current\nworkflow delta named by this contract is `ultra`'s proactive guidance; it does\nnot prove that provider reasoning itself is deeper or better.\n\n## Persistent Writes On Explicit Change\n\nDo not claim that a saved-file edit changed the current interactive TUI. An\nexplicit request such as \"use xhigh\", \"turn summaries off\", or \"set verbose\noutput to less\" authorizes the corresponding persistent setting edit when the\nrequest is unambiguous. If the requested value is ambiguous, ask which saved\nvalue to use and STOP without a mutation. A clarification timeout,\nauto-cancellation, dismissal, empty reply, or missing reply is not authorization:\nleave the setting unchanged. Never pick a value because it seems more common,\nlikely, stronger, or closer to a default.\n\nUse these stable key contracts:\n\n- `reasoning_effort` is the top-level persistent effort tier. Meta accepts\n  `minimal`, `low`, `medium`, `high`, `xhigh`, `max`, or `ultra`.\n- `tui.reasoning_summaries` is a persistent boolean; absent defaults to `true`.\n- `tui.verbose_output` is one of `less`, `edits`, or `more`; absent defaults to\n  `edits` (shown as `edits & writes`).\n- `tui.terminal_background` is one of `auto`, `light`, or `dark`; absent\n  defaults to `auto` (detect from the startup probe). `light`\u002F`dark` override\n  the detected terminal background for theme resolution at the next startup.\n\nWhen a write is requested:\n\n1. Read the current settings first.\n2. Make the smallest JSON change that satisfies the request. If the read proves\n   the file is missing, create a minimal object containing `schema_version: 1`\n   and only the requested setting. A denied or malformed file is not a missing\n   file and MUST NOT be replaced.\n3. Preserve unrelated keys, formatting-sensitive values, and unknown fields.\n4. Re-read the file to prove the saved value changed. Do not treat the reread as\n   live\u002Feffective-state evidence.\n5. Report the old saved value (or absent plus its effective default), the new\n   saved value, that it is permanent, that the current session is unchanged,\n   and that it takes effect on the next Muse Code launch. Relaunch Muse Code; do\n   not tell the user to restart the terminal application.\n\nFor a YOLO shell fallback, an earlier successful shell read does not replace the\nrequired `edit_file` or `write_file` attempt. Try the file mutation tool on the\nsame path first; use the atomic structured shell mutation only after that tool's\nexplicit denial in this turn.\n\nIf the allowed access path above still cannot read or write the active config,\nreport that no saved setting changed and stop. Do not probe another config root\nor bypass a security control.\n\n## Completion Report\n\nFor read-only answers, include:\n\n- the settings path checked;\n- the key path or default that answers the question;\n- whether any file was changed.\n\nFor a verified write, use this user-facing summary shape:\n\n```text\nSaved changes\n- \u003Ckey>: \u003Cold saved value or absent + default> -> \u003Cnew saved value>\n\nPermanent: yes\nCurrent session: unchanged\nEffective: next Muse Code launch\nReload: required - relaunch Muse Code; no terminal-application restart\nVerified: `settings.json` reread at \u003Cpath>\n```\n\nFor a setting such as `reasoning_effort` with higher-precedence startup input,\nappend `unless a launch flag overrides the saved value` to the\nEffective line. Name any setting intentionally left unchanged.\n\nIf the read, write, or verification fails, use:\n\n```text\nChanged: no\nReason: \u003Cexact path and failure>\nReload: not applicable\n```\n",{"name":5,"description":7},"main","2975a5ac469614541ae22929d43b6da2635e1ee9","Meta\u002Fmuse-code\u002Fskills\u002Fmanage-settings","https:\u002F\u002Fgithub.com\u002FraaOS\u002Fsystem_prompts_leaks","https:\u002F\u002Fgithub.com\u002FraaOS","2026-11-14 21:19:17",{"source":8,"skillName":5,"commands":30,"prompt":40,"zipAvailable":41},[31,34,37],{"tool":32,"command":33},"npx","npx skills add raaOS\u002Fsystem_prompts_leaks --skill manage-settings",{"tool":35,"command":36},"bunx","bunx skills add raaOS\u002Fsystem_prompts_leaks --skill manage-settings",{"tool":38,"command":39},"pnpm","pnpm dlx skills add raaOS\u002Fsystem_prompts_leaks --skill manage-settings","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FraaOS\u002Fsystem_prompts_leaks\u002Fblob\u002Fmain\u002FMeta\u002Fmuse-code\u002Fskills\u002Fmanage-settings\u002FSKILL.md",true,[],[44,50,57,64,71,78],{"skillNo":45,"name":46,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":47,"categories":48,"keywords":49,"publishedAt":18},"01M77JV1M0ZZTZZREM9Q96TKZZ","model-council","https:\u002F\u002Fgithub.com\u002FraaOS\u002Fsystem_prompts_leaks\u002Fblob\u002Fmain\u002FPerplexity\u002Fskills\u002Fmodel-council\u002FSKILL.md",[],[],{"skillNo":51,"name":52,"nameZh":6,"description":53,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":54,"categories":55,"keywords":56,"publishedAt":18},"01M77JTYND3ETS9EGCX3B9CW6E","writing-style","Draft or edit writing in the user's voice for the person who will read it and where it will appear. Use for messages, emails, documents or slides written on their behalf, not for your own replies.","https:\u002F\u002Fgithub.com\u002FraaOS\u002Fsystem_prompts_leaks\u002Fblob\u002Fmain\u002FOpenAI\u002Fdots\u002Fskills\u002Fwriting-style\u002FSKILL.md",[],[],{"skillNo":58,"name":59,"nameZh":6,"description":60,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":61,"categories":62,"keywords":63,"publishedAt":18},"01M77JTVWEHY80JC2H1JG1JMQB","write-page","Create or edit requested Page\u002FSpace content, or use for prose you have already decided to save as a standalone Markdown file, only if the user did not explicitly request a Markdown file. Honor established formats, destinations, and repository documentation. A selected Page alone does not authorize a write.","https:\u002F\u002Fgithub.com\u002FraaOS\u002Fsystem_prompts_leaks\u002Fblob\u002Fmain\u002FOpenAI\u002Fdots\u002Fskills\u002Fwrite-page\u002FSKILL.md",[],[],{"skillNo":65,"name":66,"nameZh":6,"description":67,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":68,"categories":69,"keywords":70,"publishedAt":18},"01M77JTQ5ND23QPF6RJYJPXHSZ","software-engineering","For use only in threads for this proactive personal assistant. Investigate software issues; inspect, write, review, or test code; work on a repository or local engineering file; or create, fix, or monitor a pull request or CI run. Use for actual engineering or product QA work, not a general programming explanation.","https:\u002F\u002Fgithub.com\u002FraaOS\u002Fsystem_prompts_leaks\u002Fblob\u002Fmain\u002FOpenAI\u002Fdots\u002Fskills\u002Fsoftware-engineering\u002FSKILL.md",[],[],{"skillNo":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":18},"01M77JTM3ZDPCYE8C1FW7G7S2B","slides-artifact","Use when the user asks to create, edit, or review slides or a presentation (PowerPoint or Google Slides), including when the task clearly calls for an audience-ready slide deck. Also use to answer questions that require inspecting an existing deck; not for general presentation advice or a chat-only outline.","https:\u002F\u002Fgithub.com\u002FraaOS\u002Fsystem_prompts_leaks\u002Fblob\u002Fmain\u002FOpenAI\u002Fdots\u002Fskills\u002Fslides-artifact\u002FSKILL.md",[],[],{"skillNo":79,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":84,"publishedAt":18},"01M77JTHTWZN2J27KD2H66XDBZ","slack","Use when reading Slack conversations, deciding when to respond, or sending messages, files, and reactions in Slack.","https:\u002F\u002Fgithub.com\u002FraaOS\u002Fsystem_prompts_leaks\u002Fblob\u002Fmain\u002FOpenAI\u002Fdots\u002Fskills\u002Fslack\u002FSKILL.md",[],[]]